Update README.md

Use validated Build Small tag slugs (track/sponsor/achievement)

Add field notes (lessons learned) to README and FIELD_NOTES.md

Add demo video link to README

Warm boss card art in battle (GPU free post-draft) so boss plays show art instantly

Pre-bake backbone card art (instant, no shimmer) + stop warming face-down enemy hand

Speed up boss deck (parallel picks, ~22s->7.5s) and forge it from run start; deck-aware README

Polish demo: boss reveal screen, portraits, varied packs, paced draft, title copy

Fix card truncation: lead JSON with effects+name, request concise flavor/art, bump card tokens

Trim prefetch to 1 pack, log model-pack failures, JSON-constrained cards, rotate fallback names

Size Modal pools under the 10-GPU plan cap (card 3/4, art 3/4, boss 1/2)

Salvage card effects from raw model text; min_containers=5 for demo

Speed: reveal first pack immediately, defer boss deck, more Modal containers

Hardcode Modal endpoint URLs in app_hf.py (public; env still overrides)

Pin diffusers==0.31.0 on Modal art image (transformers 4.49 compat; avoids flux2/Qwen3 import)

Add MODE switch: LOCAL (on-device, default) vs MODAL (GPU endpoints)

Modal-backed Space: run models on Modal GPU endpoints, Space is a thin HTTP client

ZeroGPU: move models to CUDA inside @gpu calls, not in the main process

ZeroGPU: read models from module globals, not pickled args

Pin transformers for MiniCPM-V-4; expose module-level demo for HF Spaces

Integrate ZeroGPU Space path with card-cache, sampling, and diversity fixes

Update README

Prepare Space deployment flow

Update public README

finished base

Build Tabras card duel prototype

Update README.md

Initial commit